Start here
New here? Start with these.
I'm Munir, a Lead Engineer in Sydney. I write about resilient systems, delivering safely with AI in regulated environments, and leading the engineers who do both. If you only have a few minutes, start with these.
Two AI agents deleted production databases. Neither was stopped by its instructions. Approval improves an agent's decisions; it doesn't limit what the agent can do. Why agent safety is a permissions and architecture problem first, how to size the risk of an action, and what humans are actually for.
Resilient systems
Most of my best engineering work starts with one question: what happens when this goes wrong?
Safe AI delivery
Using AI agents where mistakes are expensive: what to delegate, what to keep, and what stops an agent when it’s wrong.
Engineering leadership
Decisions, trade-offs and the part of the job that isn’t code.
Want more? Browse Systems, AI Delivery or Leadership, read my notes, see the Bookshelf, or read about me.